DICION PAPER ERIE - IZA Institute of Labor Economicsftp.iza.org/dp11926.pdfIA DP No. 11926 OCTOBER 201 Unlucky Cohorts: Estimating the Long-term Effects of Entering the Labor Market

DISCUSSION PAPER SERIES

IZA DP No. 11926

Hannes SchwandtTill von Wachter

Unlucky Cohorts: Estimating the Long-term Effects of Entering the Labor Market in a Recession in Large Cross-sectional Data Sets

OCTOBER 2018

Any opinions expressed in this paper are those of the author(s) and not those of IZA. Research published in this series may include views on policy, but IZA takes no institutional policy positions. The IZA research network is committed to the IZA Guiding Principles of Research Integrity.The IZA Institute of Labor Economics is an independent economic research institute that conducts research in labor economics and offers evidence-based policy advice on labor market issues. Supported by the Deutsche Post Foundation, IZA runs the world’s largest network of economists, whose research aims to provide answers to the global labor market challenges of our time. Our key objective is to build bridges between academic research, policymakers and society.IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion. Citation of such a paper should account for its provisional character. A revised version may be available directly from the author.

Schaumburg-Lippe-Straße 5–953113 Bonn, Germany

Phone: +49-228-3894-0Email: [email protected] www.iza.org

IZA – Institute of Labor Economics

DISCUSSION PAPER SERIES

IZA DP No. 11926

Unlucky Cohorts: Estimating the Long-term Effects of Entering the Labor Market in a Recession in Large Cross-sectional Data Sets

OCTOBER 2018

Hannes SchwandtNorthwestern University, SIEPR, CEPR and IZA

Till von WachterUniversity of California Los Angeles, NBER and CEPR

ABSTRACT

IZA DP No. 11926 OCTOBER 2018

Unlucky Cohorts: Estimating the Long-term Effects of Entering the Labor Market in a Recession in Large Cross-sectional Data Sets1

This paper studies the differential persistent effects of initial economic conditions for labor

market entrants in the United States from 1976 to 2015 by education, gender, and race

using labor force survey data. We find persistent earnings and wage reductions especially

for less advantaged entrants that increases in government support only partly offset. We

confirm the results are unaffected by selective migration and labor market entry by also

using a double-weighted average unemployment rate at labor market entry for each birth

cohort and state-of-birth cell based on average state migration rates and average cohort

education rates from Census data.

JEL Classification: J2, J3, J6

Keywords: job market entry, recessions, unemployment, long-term effects

Corresponding author:Hannes SchwandtSIEPRStanford University366 Galvez Street, StanfordCA 94305United States

E-mail: [email protected]

1 We thank Ioannis Kospentaris and Nicolas Oderbolz for excellent research assistance, and David Card, Janet

Currie, Lisa Kahn, Adriana Lleras-Muney, and participants at the NBER Youth conference and several seminars for

helpful comments.

1 Introduction

The first years after entering the labor market are typically a very productive period for youngworkers. During this period, young workers’ wages grow rapidly and they frequently switchtowards better paying jobs.2 At the same time, young workers are particularly vulnerable toadverse conditions in the labor market. For example, it is well known that young workersbear the brunt of recessions in terms of higher unemployment rates, partly because wagestend to fall most for those workers entering new jobs (e.g., Elsby et al. (2016)). Economistsand policy makers alike have long been concerned that interruptions of the initial process ofcareer progression caused by recessions can have lasting consequences on earnings and otherrelevant outcomes, including health insurance coverage, health effects, and family formation(e.g., Von Wachter (2012)).

There is indeed increasing evidence from careful studies of college graduates that eventemporary exposure to increased unemployment rates can lead to persistent earnings reduc-tions. For example, using data from the National Longitudinal Study of Youth (NLSY),Kahn (2010) shows that college graduates entering the labor market during the deep reces-sion in the early 1980s experienced reductions in earnings lasting up to fifteen years. Oyer(2008; 2006) presents evidence on the persistent effect on career choice for MBAs and PhDEconomists. Oreopoulos et al. (2012) show that college graduates in Canada suffer persistentearnings losses and that these losses are substantially larger for those graduates predicted tohave low earnings to begin with.3 Outcomes other than earnings appear to respond to initiallabor market entry as well.4 For example, based on the NLSY, Maclean (2013) finds thatmale entrants during the early 1980s recession have experienced long-lasting effects on self-reported health, which is consistent with findings based on mature workers that show labormarket shocks can have long-term effects on health, including mortality.5

These studies provide powerful evidence that the fear that recessions can have lastingrepercussions for young workers is well founded. These findings are worrisome, in partic-ular because new college graduates are typically not eligible for programs meant to buffer

2E.g., Topel and Ward (1992), Murphy and Welch (1990).3See also Altonji et al. (2016) for analysis of differential effects of graduating in a recession by college

majors.4E.g., Altonji et al. (2016) analyze earnings and occupational choice of college graduates using multiple

data sources, spanning a similar time period as we do; Giuliano and Spilimbergo (2014) analyze the effects onattitudes; Oreopoulos et al. (2012) study employer characteristics.

5E.g., Sullivan and Von Wachter (2009) show that mature job losers suffer long-term increases in mortalityrates. We explore the effect of entering the labor market in a recession on long-term mortality in a companionpaper (Schwandt and Von Wachter (2018)).

1

temporary earnings losses. By focusing on college graduates, the aforementioned papershave been able to provide empirical evidence that is highly compelling and provides impor-tant proof-of-concept results establishing the significance of persistent effects of early labormarket conditions. This is partly because they can exploit high-quality longitudinal datathat allow measuring the state in which workers first entered the labor market, among othersthings, and partly because the date of entry and typical career progression is well defined forcollege graduates.

However, it is well known that less-advantaged groups in the labor market, such as low-educated workers or minorities, experience much larger increases in unemployment duringrecessions (e.g., Hoynes et al. (2012)). These groups are thus at risk of suffering even largerlonger-term effects than the highly educated workers studied in depth in the existing liter-ature. At the same time, in contrast to college graduates, these workers are more likely tohave access to the social safety net. Indeed, recent work from European countries suggeststhat entry conditions have a stronger effect on a range of outcomes, including self-reportedhealth, for lower-educated individuals (e.g., Cutler et al. (2015)), despite the wide-spreadprevalence of generous social support systems in these counties. Despite the concern thatless advantaged individuals could fare worse, the effect of adverse labor market entry forthese groups of workers has not been studied extensively in the literature, especially for theUnited States, where safety nets are less extensive. This is partly because typical longitu-dinal data do not have sufficient samples to study these groups and partly because careerprogression, and hence initial conditions, are harder to measure.6

To advance the literature, in this paper we examine the persistent effects of entering thelabor market in a recession on a broad range of socio-economic outcomes for all youngworkers who entered the labor market in the United States from 1976 to 2015. Our study in-cludes college graduates, but also focuses on groups not typically analyzed separately, suchas women, non-whites, and individuals with less than a college degree. To identify the effectof initial labor market conditions, we exploit year-to-year variation in unemployment ratesin the state of labor market entry. To estimate these effects, we use several data sets with ex-tensive coverage over time and large sample sizes: repeated cross-sections from the AnnualSocial and Economic Supplement to the Current Population Survey; Decennial Census data

6In an exception, Speer (2016) uses the NLSY to study the effects of initial labor market conditions forlower educated men graduating chiefly in the early 1980s recession and finds similar results to ours. There isa separate body of literature on the scarring effects of individual labor market shocks for young workers, suchas unemployment spells, occurring independently of macroeconomic conditions (e.g., Gardecki and Neumark(1998)). While identification is difficult, Von Wachter and Bender (2006) pursue an instrumental variablesapproach based on temporary condition at workers’ first employer.

2

and American Community Survey data.Given that these data sets contain information on a large number of entry cohorts in

the labor market, they allow us to generate the first estimates of entering the labor marketin a recession for a typical young labor market entrant in the United States. Another keyadvantage is that the large sample sizes allow us to study with sufficient precision the effectfor smaller groups, such as high school dropouts. Finally, the data contain information ona range of additional outcomes that have not been studied for young workers in the U.S.labor market. This includes information about the role of the social insurance system foryoung workers, such as receipt of Medicaid and Supplemental Nutrition Assistance Program(SNAP, formerly Food Stamps), as well as measures of poverty.

These substantive advantages come at a price in terms of precision of our research designimposed on us by the data. In particular, the cross-sectional data we use do not containinformation on the timing and location of entry into the labor market. The first data issueconcerns potential endogenous timing of labor market entry, something all studies of thiskind have to deal with.7 The second is unique to our use of cross-sectional data becauseregional mobility can introduce either random measurement error or systematic bias. Giventhe importance of these measurement aspects, we address these issues head on in the paperusing several approaches. Overall, after careful analysis, we conclude that our approach forstudying the persistent effect of adverse labor market conditions based on repeated cross-sectional data is feasible and yields very similar findings to estimates that explicitly correctfor mobility or endogenous labor market entry.

Based on this approach, we obtain three key findings. First, for the full sample of labormarket entrants in the United States from 1976 to 2015, we find that entering the labor mar-ket in times of high unemployment leads to a substantial initial effect on earnings. Consistentwith findings in the previous literature, this effect fades gradually, but persists for ten yearsinto workers’ careers. Our findings imply that for a moderate recession that raises unem-ployment rates by three points, the loss on cumulated earnings is predicted to be on the orderof 60% of a year of earnings. These effects are substantial and very robust to controls forselective migration or endogenous entry into the labor market. Further analysis suggests thatthe initial effect is due to both employment and wage reductions, whereas the longer-termeffect is mainly due to persistent declines in wages.

Second, we find that the effect on earnings varies considerably in the population. Whileall groups we studied experience persistent effects from adverse initial labor market condi-

7However, our analysis of lower-education groups has to deal with the possibility that our approach inap-propriately includes individuals whose education is in progress. We discuss this explicitly in the paper.

3

tions—including women, high school graduates, and those with some college—the effectsare particularly large for two groups: nonwhites and high school dropouts. Although smallersamples lower the precision for these groups, the earnings losses are substantial. In examin-ing the sources, we find these differences are partly driven by greater losses in employment,measured in terms of the number of weeks worked in the past year, for nonwhites and highschool dropouts.

Third, we find that the U.S. social insurance system provides a buffer for unlucky labormarket entrants and that these effects are largest for those who suffer the greatest earningslosses. We find precisely estimated temporary but long lasting increases in the probabilityof receiving Supplemental Nutrition Program Assistance (SNAP, formerly known as FoodStamps) for the full sample. The effects are present for both men and women and whites andnonwhites and are driven by a rise in probability of receiving benefits among those with ahigh school degree and high school dropouts. As a result, the effect on household income– our measure of income available from all sources – is lower than the effect on annualearnings. However, the insurance provided is imperfect, and we find effects on povertylasting up to six years among all groups except those with at least some college or more.

We also find that adverse initial labor market conditions raise the receipt of Medicaidfor all groups with exception of those with at least some college or more. Again, the ef-fects are particularly strong for nonwhites, and high school dropouts. For those with a highschool degree, a rise in the probability of receiving Medicaid appears to buffer a temporary,employment-related loss in private health insurance coverage. Only those with some collegeexperience a temporary reduction in private health insurance coverage that leads to a net de-cline in any health insurance receipt, but the effect is short lived. College graduates do notexperience a reduction in health insurance coverage.

These findings extend the literature on persistent effects of temporary labor market con-ditions along several dimensions. The foregoing literature, especially on U.S. youth, con-centrated mainly on college graduates. Although there are some findings based on broadersamples, this is the first study to comprehensively address and compare differences in thepersistent effect of initial labor market conditions for labor market entrants in the UnitedStates.

Our results also provide useful information regarding the importance and effect of thesocial insurance system in buffering cyclical employment effects. Most of the focus in thisarea is on mature workers, and in particular what sources of income are available for thelong-term unemployed (e.g., Rothstein and Valletta (2017)). But little is known about how

4

the social insurance system helps labor market entrants weather weak economic conditions.Young workers are typically not covered by unemployment insurance and are usually single,which excludes them from typical welfare programs, such as Temporary Aid for NeedyFamily.

Beyond our substantive contributions to this literature, our paper provides a novel method-ological approach with two key advantages that will be helpful to researchers studying thisphenomenon who seek to go beyond smaller longitudinal data sets that may not have suffi-cient precision for this type of analysis. First, our approach allows bias from selection andmeasurement error due to endogenous migration and education decisions. Second, it allowsfor the harnessing of much larger cross-sectional data sets as long as information on state ofbirth is used. The approach constructs the relevant unemployment rate for each birth cohortand state-of-birth cell on which our main analysis is based by properly weighting and ag-gregating current state unemployment rates using average state migration rates and averagecohort education rates.

Finally, our results provide a useful practical contribution to the literature by analyzingpersistent effects of the local environment during youth and early adulthood. These effects,among others, have received recent attention in studies of intergenerational transmission ofincome and the role of neighborhood effects based on rich longitudinal data (e.g., Chettyet al. 2016; Chetty and Hendren 2018).8 Our findings suggest that labor market mobilitymay be sufficiently low and idiosyncratic that the state of residence after entry into the labormarket (or state of birth) approximates the characteristics of the initial location at the timeof entry into the labor market sufficiently well on average. These findings confirm earlierresults of Card and Krueger (1992), which find mobility adjustments do not affect results ofthe earnings effects of school characteristics, as well as the evidence from Autor et al. (2014),which shows that local trade shocks do not lead to significant migration to less affected areas.

The remainder of the paper is structured as follows. Section 2 describes our empiricalapproach, our data, and how we assess whether the cross-sectional data can be successfullyused to estimate the long-term effect of initial conditions. Section 3 summarizes the effectof initial unemployment rates on the socio-economic outcomes we study, including annualearnings, hourly wages, employment, program receipt, and health insurance coverage. Sec-tion 4 concludes.

8Another related literature analyzes the effects of local economic conditions on fertility (Currie andSchwandt (2014)).

5

2 Empirical Approach and Data

We seek to extend the existing literature that focuses on college graduates by studying theeffect of entering the labor market in a recession for more disadvantaged groups in the labormarket. To do so, we use data from repeated cross-sections in the March Current PopulationSurvey (CPS), the Decennial Census, and the American Community Survey. This approachhas several advantages in our context. It allows us to work with much larger samples andhence enables us to study the responses of smaller subgroups, such as nonwhites or low-educated workers. The data cover a longer time period thereby allowing us to analyze theeffects of entering the labor market for all graduating cohorts from 1976 to 2015. This isthe first paper to do so in the United States, and it is only possible due to the use of cross-sectional data. In addition, information in the March CPS data allow us to analyze additionaloutcomes that are particularly relevant for lower-income workers.

Working with cross-sectional data has drawbacks as well, all of which we address di-rectly. An important limitation is that we neither know the actual state nor the exact timingof entry into the labor market. Moreover, both state and time of labor market entry might beendogenous if people respond to local recessions by moving into other states or by postpon-ing graduation. In what follows, we will start with a hypothetical ideal regression that doesnot suffer from endogeneity or measurement issues. In Section 2.2, we discuss how one canproxy for the state and year of graduation in the data given this ideal empirical model, andwhich biases may arise from this specification. In Section 2.3, we develop an approach thataccounts for these biases in the Census and ACS data.

2.1 The Ideal Regression

The ideal regression would relate different outcomes such as labor income at different yearsof experience to the economic conditions (eci0) an individual i faced at her or his labormarket entry.

yi,t = ↵ + �eeci0 + �e + ✏i,t (1)

where �e are experience fixed effects. The coefficients �e represent deviations from thetypical experience profile resulting from differences in local labor market conditions. Acausal interpretation of the coefficient estimates for �e requires that the economic conditionsat labor market entry are uncorrelated with other determinants of the respective outcome.This equation can be derived from economic models relating wages and employment to

6

local labor market conditions (e.g., Blanchflower and Oswald (1994)). To implement theequation, we have to choose a level aggregation for the appropriate labor market. For ouranalysis, we follow the majority of the literature and use annual state-level unemploymentrates.

2.2 Mincerian Specification

Our main analysis is based on CPS data that provide detailed individual-level informationon labor market outcomes. However, the CPS does not report the year when an individualcompleted her education, nor the state where she entered the labor market. We use as aproxy the “Mincerian” graduation year, i.e. the sum of the year of birth, plus 6, plus theyears of reported education. The state of labor market entry is proxied by the state of currentresidence, which is the only state identifier reported in the CPS.

For our baseline specification, we follow the literature (e.g., Oreopoulos et al. (2012)(henceforth OWH)) and work with a cell-based model that aggregates the outcome at thelevel of current state of residence (s), year of graduation (g), calendar year (t), and educationgroups (d). Working with the cell-level data is sufficient because we do not use controlvariables varying at the individual level, and it allows us to work close to our main source ofvariation coming from local unemployment rates. We regress the average, cell-level outcomeon the relevant initial unemployment rate, and control for year of graduation, state, and yearfixed effects:

ȳs,g,t,d = ↵ + �eus,g + �e + �s + �g + ✓t + ⇡d + ✏s,g,t (2)

where us,g is the unemployment rate in the state of current residence s at the “Mincerian”year of graduation g.9 e refers to years of potential experience (years since graduation) andt to the calendar year. We additionally include education group fixed effects ⇡d. In contrastto previous work, our analysis includes all educational groups in one specification given thatstate-cohort-level variation in educational attainment could be a confounding factor. Noticethat we do not include the current state unemployment rate in our main results; therefore�e captures the effect of graduating in a recession, given the regular subsequent evolutionof the local labor market conditions.10 To properly represent population-level relationships,

9For ease of reference, in some cases we will refer to the implied year of entry as a graduation cohort (eventhough in some cases individuals do not literally graduate), and to the unemployment rate in the implied yearof entry as ’graduation unemployment rate’. Similarly, if we refer to ’state’ without further clarification, thestate of current residence is intended.

10The effect of the initial unemployment rate consists of its own direct effect, plus the weighted effect of

7

the cell-level observations are weighted by the corresponding cell sizes. Standard errorsare clustered at the level of graduation year by state to account for cohort-specific serialcorrelation in labor market outcomes.

Given the included fixed effects, the coefficient vector �e captures deviations from thetypical experience profiles related to cohort-state-specific variation in the unemploymentrate at labor market entry that are uncorrelated with contemporaneous nation-wide shocks.However, the specification does not account for cohort-state specific variation driven byendogenous graduation timing and migration that might bias our estimates.

Endogenous graduation timing. Our baseline specification treats the time of labormarket entry (proxied by years of education plus 6) as exogenous. But people might pro-long their educational attainment in order to avoid unfavorable conditions at entry or endtheir education prematurely in order to benefit from good labor market conditions. Suchendogenous timing of labor market entry attenuates our estimates towards zero if it is uni-formly distributed among new labor market entrants. If there is selection into timing, the biascan go either way. For example, if those with higher potential earnings are better in timingtheir labor market entry, then we would tend to overstate the effects of initial labor marketconditions.11

Endogenous migration before and after graduation. In the CPS data we proxy forthe state of graduation using the current state of residence. However, in response to a localrecession around the time of labor market entry, people might migrate into other states thatare less affected. Such directed migration would lead to an attenuation bias in our data, as themigrants from poorly performing states would be erroneously assigned the better economicconditions in their new state of residence. If there is selection in who tends to leave inresponse to adverse economic conditions, the bias could go either way. We tested for suchselection effects with balancing regressions that use the racial or gender composition as adependent variable in our Web Appendix (Section IV), and found numerically small effects(Pei et al., 2018).

Undirected migration after graduation. People might migrate independently of lo-cal labor market circumstances (“undirected” migration). The implied mismeasurement of

subsequent unemployment rates correlated with it (see OWH for a more detailed discussion). In comparing twocohorts with different initial conditions, this captures the full difference in life-time earnings due to adverselabor market entry. Results controlling for the current unemployment rate are shown in the Web Appendix,Section XI.

11We can test for the presence of endogenous timing by regressing a cohort’s share of high school and collegegraduates on the unemployment rate at age 18. Selection into timing can additionally be explored by lookingat the racial or gender composition of these graduation cohorts. These results are shown in the Web Appendix(Section III and Section IV, respectively).

8

initial labor market conditions would lead to attenuation bias, too, though it would be lessstrong than in the case of endogenous migration. However, in both cases the bias worsensover time as the share of migrants accumulates within graduation cohorts and the currentstate of residence becomes an increasingly poor proxy for the state of graduation.

2.3 Double-Weighted Specification

We explore the role of these potential biases in the Decennial Census (henceforth “Census”)and ACS data by comparing the Mincerian specification we estimate based on CPS data toa double-weighted specification that is not affected by endogenous timing or migration. TheCensus and ACS do not only report the state of residence but also individuals’ state of birth.We use this information to construct a proxy for the graduation-year unemployment rate thatrelies on individuals’ state of birth and year of birth. Because these characteristics are fixed,they cannot be affected by labor market conditions around graduation.12

In particular, we separately estimate average migration rates at different ages and educa-tion shares, which are then use to construct a double-weighted average unemployment ratefor each graduation year.13 This double-weighted measure provides a proxy for the typicalexposure of a cohort to economic conditions across the United States and across differentpotential graduation years that is independent of a cohort’s actual migration or graduationtiming.14

To implement this approach, we first estimate migration shares mAb,s as the average shareof cohorts in our sample born in state b that live in state s at ages A =16, 18, 20, 22. Note thatwe only use state-specific average migration rates (i.e., state fixed effects in the underlyingregression model of individual-level migration indicators) rather than migration rates of aspecific birth cohort (state-cohort fixed effects) which could be driven by an endogenousresponse to contemporaneous labor market conditions. Next, we estimate average graduationshares eAb,c, indicating the share of sample cohorts born in state b in year c who graduate at age

12This approach is based on synthetic cohorts (Deaton, 1985) and a further development of Currie andSchwandt (2014), who link maternal life cycles to unemployment rates using mothers’ own state and year ofbirth.

13The underlying assumption is that migration rates are similar across education groups. While this is notborne out in typical studies of migration, given low average migration rates it makes little difference in ouranalysis.

14To account for endogenous migration, we could simply match people to the graduation year unemploymentrate in their state of birth. However, this would imply a mismatch for those who migrated before graduation,leading to attenuation bias. For example, in the 2000 Census, about 20% of 18-year-olds live outside theirstate of birth. This attenuation is likely to be more dramatic than the one caused by random migration aftergraduation because migration rates are much higher during the first two decades of people’s lives than duringthe following two.

9

A =16, 18, 20, 22+. To predict graduation shares, we regress cohort-state-specific shares onstate fixed effects and country-wide cohort fixed effects. We then obtain eAb,c by adding stateand cohort fixed effects, such that, again, we take out the potentially endogenous cohort-by-state variation. The double-weighted (DW) average graduation year unemployment rate isthen given by:15

uDWb,c =X

A

eAb,c

50X

s=1

mAb,sus,c+A (3)

Because we are averaging both across locations and graduation years, this adjustmentprocedure reduces the amount of variation available for the estimation of our effects of inter-est. As a consequence, coefficients based on the double-weighted unemployment rate will beestimated less precisely. More important, however, such estimates will be free of potentialbias.

Endogenous migration or timing in response to a local recession are not contained inthe double-weighted unemployment rate, as it is constructed using only state averages overtime and national averages across periods, but not state-period variation. Moreover, becausethe cohorts are assigned their graduation state based on their state of birth—a fixed char-acteristic— the double-weighted unemployment rate is not affected by state-cohort-specificmigration after graduation.16

We regress earnings and income in the Census and ACS data on the double-weightedunemployment rate using the following specification:

ȳb,c,a = ↵ + �auDWb,c + �a + �b + �c + ✓t + ✏c,b,a (4)

Because the double-weighted unemployment rate is predicted at the cohort level, thisapproach requires collapsing the data by state and year of birth instead of state and year ofgraduation. This means we restrict our sample to U.S. natives and track effects over cohorts’age rather than experience profiles. The indices b, c, a, and t refer to the birth state, birth

15For example, assume cohorts born in California have a typical outmigration to Nevada of 20% by age 18and of 25% by age 22. Also assume that 60% get 12 years of education, while the remaining 40% getting 16years of education. High-school and college graduates for, say, the 1980 California-born cohort would enterthe labor market in 1998 and 2002, respectively. The double-weighted average graduation-year unemploymentrate for the 1980 California-born cohort would then be: 0.8 ⇤ 0.6 ⇤ uCA1998 + 0.2 ⇤ 0.6 ⇤ uNV1998 + 0.75 ⇤ 0.4 ⇤uCA2002 + 0.25 ⇤ 0.4 ⇤ uNV2002.

16In contrast, the actual mean unemployment rate at entry into the labor market for a given birth cohort(c) and state of birth (b) would depend on cohort-specific mobility and graduation rates (indicated by anasterisk), potentially leading to biases if migration responds to initial economic conditions: E(us,g|b, c) =P

A e⇤Ab,c

P50s=1 m

⇤Ab,s,cus,c+A .

10

year, age, and calendar year; hence �, �, � and ✓ are the coefficients on a full set of age, birthstate, birth cohort and calendar year fixed effects, respectively.

Our main approach is to compare the results of our main specification in Equation (2)based on the CPS and the results in Model (4) based on the Decennial Census and ACS. Ifthe results are similar, this indicates that migration and timing of graduation are not problemsin our sample, and we proceed with (2), a specification that can be used in the rich CPS data.

An alternative approach is to use the double-weighted unemployment rate as an instru-ment for the actual endogenous unemployment rate a cohort faces at graduation, proxied bythe “Mincerian” rate in Equation (2). We provide the 2SLS results from such instrumentalvariable regressions, which include the full set of fixed effects contained both in Equations(2) and (4). Since the estimates of Equation (2) are simply the “reduced form” estimates,the 2SLS effectively rescales our main results by the regression coefficients of the first stageregression of the “Mincerian” on the double-weighted unemployment rate.

Figure 1 plots the Mincerian and the double-weighted graduation year unemploymentrates across our sample cohorts for four large states. Since the double-weighted rate is con-structed at the level of birth year and birth place, there is only one double-weighted rate foreach birth cohort in a given state (thick solid red line). However, there is a large numberof different Mincerian rates for each birth cohort (blue circles), due to labor market entryat different ages and migration across states. The thin solid black line shows the averageMincerian rate for each cohort, weighted by the size of each graduation age x state cell. Thekey difference between the average Mincerian and average the double-weighted rate is thatthe latter accounts for endogenous migration and graduation timing. The fact that the twosolid lines follow each other closely suggests that the bias arising from such endogenousresponses in the simple Mincerian specification are unlikely to be very large.17

2.4 Sample Restrictions and Summary Statistics

State-level unemployment rates are available from the Bureau of Labor Statistics only since1976. Therefore we exclude individuals who graduated before 1976 when using the actualgraduation year and individuals who were born before 1960, i.e., age 16 before 1976, whenusing the double-weighted unemployment rate (which is based on the unemployment ratescohorts face at age 16 and above). The CPS data we use is from the Annual Social andEconomic Supplement (ASEC) fielded in March. Our last year of data is 2016. We confinethe CPS analysis to individuals between ages 16 and 40. In addition, we limit the analysis

17Corresponding numbers are shown in the Web Appendix, Section II.

11

to individuals with no more than 15 years of potential experience (when using the actualgraduation unemployment rate) or are no older than age 33 (when using the double-weightedunemployment rate). The sample for the double-weighted specification is further restrictedto individuals born in the United States.18,19

Table 1 presents summary statistics for our main sample and lists the variables we an-alyze. In addition to studying the effect of entering the labor market in a recession on logannual earnings, we also examine the effect of an unlucky start on a range of other factorsrelated to employment, income, and the social insurance system. In particular, we studythe effects on wages and weeks worked, as well as the receipt of support from programssuch as SNAP, unemployment insurance, and welfare, as well as the use of Medicaid andreceipt of health insurance more generally. To assess the incidence of initial labor marketshocks on resources available to individuals, we also study the effects on household income,a more comprehensive measure that should reflect both labor earnings as well as incomereceived from social insurance programs and spousal earnings. These outcomes are of par-ticular relevance for less-advantaged workers and might provide mediating factors for theeffects on other outcomes analyzed in the recent literature such as mortality (e.g., Schwandtand Von Wachter (2018)).

The upper part of Table 1 shows expected differences in earnings, hourly wages, house-hold income, and labor supply across groups. These differences help to underscore thatnonwhites and lower-educated workers in particular have a substantially different averagecareer outlook than college graduates on whom the existing literature has focused.20 Thebottom part of Table 1 confirms that these less-advantaged groups also benefit in particular

18For a more detailed description of sample construction, see the Web Appendix (Section I). The main CPS-based results include foreign-born individuals, as these are part of the US labor market. Excluding the foreignborn has no bearing on our findings (see the Web Appendix, Section V).

19Given the CPS does not contain information on timing of entry into the labor market, we infer the year ofentry into the labor market with the year in which the potential experience of an individual is equal to zero.This leaves some ambiguity in particular for young individuals without a high school degree, who might be stillattending school and have not actually entered full-time into the labor market. This should not be a problem forour main estimates, since they measure the differential effect on earnings and employment from the baselinedevelopment with potential experience. Since individuals that are still in school should not be affected, ourmain findings for the group of high school dropouts are likely to be based on individuals that have alreadyentered the labor market. To assess directly whether there may be a contribution from students holding summerjobs, whose earnings presumably could be affected by the concurrent unemployment rate, we confirmed thatour findings are similar when we exclude from the analysis children without a high school degree that had lessthan 13 weeks of employment (Web Appendix, Section IX).

20It is worth noting that compared to earnings, the level of household income for high school dropouts isinflated because some of these individuals are still living at home and household income likely refers to parentalincome. Again, this should not be a problem for our main estimates, since they measure the differential effecton earnings and employment from the baseline development with potential experience.

12

from social insurance programs such as SNAP or Medicaid. As has been documented else-where, there have been ongoing differential trends in labor market outcomes, especially byeducation groups. We control for those trends directly when we present separate estimatesby education group. Although we do not control for group-specific trends in our main spec-ification, there is no evidence that these secular, nationwide trends strongly correlate withcyclical fluctuations in local labor market conditions at graduation.

3 The Effect of Entering the Labor Market in a Recession on Socio-economic Outcomes

3.1 The Effect on Earnings for the Full Sample: Baseline Estimates and Sensitivity

3.1.1 Baseline Estimates

Figure 2 shows the effects of the initial unemployment rate on log annual earnings in thefirst 15 years in the labor market. The figure displays the coefficients �e on the interaction ofdummies for potential experience with the unemployment obtained from estimating Equation2 for the entire sample of labor market entrants. The point estimates with standard errors forfive experience groups are shown in Table 2. The results clearly show, as expected, thatearnings at labor market entry fall when the local unemployment rate rises. The effects aresubstantial: for a three-point rise in the unemployment rate—roughly the typical increasefrom peak to trough of the business cycle—the point estimates in Table 2 suggest an initialreduction of earnings by approximately 11%. This effect only slowly declines with timespent in the labor market. The reduction is still significantly different from zero 10 yearsafter graduation (a reduction of 2.6% for a three-point rise in initial unemployment rates),but then fades to zero as shown in Figure 2.

These estimates confirm previous findings mostly based on college graduates and formore narrow time periods. Our results tend to be somewhat larger than previous studies.While the estimates are difficult to compare because of differences in cohorts and time peri-ods included, the study by OWH is most comparable because it also includes a broad numberof cohorts covering multiple recessions. Focusing on college graduates, OWH find an initialearnings loss of approximately 6% for a three-point rise in the local unemployment rate thatfades over time. For U.S. college graduates during the severe 1980s recession, Kahn (2010)finds somewhat larger estimates. According to our estimates, a one-point rise in the initialunemployment rate reduces cumulated earnings by approximately 20% of average annual

13

earnings in the sample (summing the coefficients in Table 2 and scaling for the number ofexperience years they represent). Hence, a recession—roughly corresponding to a three-point rise in the unemployment rates—would lead to a reduction of cumulated earnings byapproximately 60% of average annual earnings, a substantial effect. Relative to total earn-ings in the first ten years of the labor market, the effect is approximately 6%.21 The fact thatour findings tend to be somewhat larger is likely due to the fact that we include in our anal-ysis more vulnerable groups than are typically studied, something we return to in Section3.3. Before turning to that, we first use our double-weighted unemployment rate measure toshow that these effects are not due to the fact that we are using cross-sectional data.

It is worth noting that our estimates on the effect of initial unemployment rates on annualearnings are likely to be underestimates of the actual total effect. This is because as furtherdiscussed below we find that the overall rate of employment declines temporarily due to ahigh initial unemployment rate. This leads to two sources of under-estimation of the true lossin earnings due to initial labor market conditions. Our estimates exclude workers that havezero earnings and hence do not count the losses deriving from longer spells of nonemploy-ment. In addition, given we find that negative employment effects tend to be strongest forlower-educated individuals, a model of monotone selectivity into employment based on, say,underlying skill would suggest that the resulting sample selection bias is positive. In otherwords, individuals with the lowest earnings potential are less likely to work as a result due tothe initial shocks, biasing the result towards zero. We have assessed the potential effect forour estimates based on log annual earnings by imputing a low earnings amount, confirmingthat our estimates likely understate the true loss by a moderate but non-negligible amount(see Web Appendix, Section VII).

3.1.2 Correcting for Interstate Mobility and Endogenous Labor Market Entry

In Section 2.3 we describe the biases that can arise from interstate migration and graduationtiming and how they can be corrected using the Decennial Census and ACS data. Figure3 shows how these corrections affect our baseline estimates. The red triangles show theCPS estimates for annual earnings as in Figure 2. The blue squares show that the baselinespecification results in very similar estimates in the Census/ACS data.

For the third set of estimates in Figure 3, the solid black line without markers, we userespondents’ state of birth instead of state of residence as a proxy for the state of gradua-

21These numbers are upper bounds because the fact that experience profiles are increasing implies thatpercentage losses earlier in a career receive a lower weight in an appropriately weighted total.

14

tion. As explained above, this specification is not affected by any endogenous or randommigration after graduation, but there is likely attenuation due to migration before gradua-tion. The resulting estimates are attenuated by about 20% in the first years after graduationin comparison with the baseline specification, as one would expect given a pre-graduationmigration rate of about 20%. The difference between estimates fades at higher experienceyears, as accumulative migration after graduation attenuates the baseline estimates but notthose based on the state of birth.

Finally, the hollow green markers show estimates based on the double-weighted unem-ployment rate, which corrects for migration (both before and after graduation) as well as forendogenous labor market entry. The effect profile is shown over age instead of experience,as the double-weighted unemployment rate is constructed at the level of year and state ofbirth. The adjusted effects seem a bit more noisy, likely due to the adjustment-induced re-duction in the amount of identifying variation. But despite the loss in precision, effects arevery similar to the baseline specification in the first years of age / experience plotted in thefigure. And in line with a slight attenuation of the baseline specification due to accumulativerandom migration after graduation, the adjusted effects are somewhat stronger in later years.Overall, these results suggest that in our baseline specification, any bias due to endogenoustiming of labor market entry and interstate migration is very limited. This conclusion is inline with the close correspondence of the Mincerian and the double-weighted unemploymentrate shown in Figure 1.

Figure 4 shows results from an instrumental variables (IV) specification that uses thedouble-weighted unemployment rate as an instrument for the endogenous “Mincerian” rate.The red triangles, blue squares, and hollow green markers repeat the two baseline specifi-cations and the double-weighted specification from the previous figure, respectively. Notethat the double-weighted specification represents the reduced form in the IV setting. Thefirst stage regression of the “Mincerian” on the double-weighted unemployment rate is 0.7with a t-statistic of 24. The thick solid black line in Figure 4 shows the IV estimate and itis very similar to the other specifications. Our main takeaway is that these different specifi-cations can be used interchangeably given the limited role of endogenous graduation timingand interstate migration. With this in mind, we return to the CPS analysis that is based onthe “Mincerian” specification.

15

3.2 The Effect on Other Outcomes for the Full Sample

3.2.1 Effects on Weeks Worked and Wages

The CPS data allow us to decompose the earnings effect into an effect stemming from areduction in the number of annual weeks worked, a reduction in usual hours worked perweek, and a reduction in hourly wages (calculated by dividing total annual earnings last yearby approximate total annual hours in the survey week, a common procedure). The results,shown in Figure 5, indicate some interesting patterns. First, exposure to high unemploymentrates at labor market entry leads to a precisely estimated persistent reduction in hourly wageslasting all 15 experience years included in the analysis. While the effect after 10 years inthe labor market is small, clearly an unlucky initial start depresses earnings even for thoseindividuals obtaining a job. Given that these estimates are based on a potentially positivelyselected group of individuals who found jobs, they may understate the true reduction inearnings capacity for unlucky entrants. Second, we find non-negligible effects on weeksworked that are concentrated in the first five years after labor market entry. Finally, we finda smaller but surprisingly persistent reduction in usual hours worked. An examination of thepoint estimates shown in Table 2 show that about two-thirds of the effect on annual earningswe find in the first three years is driven by a reduction in total hours worked (weeks workedtimes usual hours). This drops to 50% in experience years 4 to 5. In contrast, two-thirds ofthe longer-term effect of adverse initial labor market entry on annual earnings is driven by areduction in hourly wages.

It is worth noting that these findings are based on using the log of weeks worked last yearas an outcome, and hence do not count the loss in labor supply arising because of nonem-ployment spells lasting the entire survey year. We have separately analyzed the incidence ofzero employment in the survey year (i.e., zero weeks worked). For added context, we alsocompared it to the incidence of employment, unemployment, and rate of being out of thelabor force during the survey week. The results, shown in the Web Appendix (Section VI),confirm that entering the labor market when local unemployment rates are high depressesemployment for about 5 to 6 years, including a rise in the incidence of longer-term nonem-ployment spells.22

22Interestingly, for all of our most of these longer spells appear to be related to exits from the labor forcerather than unemployment. This is not surprising given self-reported unemployment is often associated withreceipt of unemployment insurance benefits, for which many unlucky young labor market participants do notyet qualify because of a lack of earnings history.

16

3.2.2 Effects on Social Welfare and Total Income

Figure 2 shows the effect of a high initial unemployment rates on the log of family income.There is a clearly visible negative and persistent effect of initial labor market conditions thatis precisely estimated even ten years into the labor market. However, the impact is smallerthan for annual earnings, especially in the first years after labor market entry. A cumulationof the coefficients in Table 2 implies that the lifetime effect of a three-point rise in the initialunemployment rate is roughly 30%, about half the effect of earnings discussed in Section3.1.1.

Transfers from the social insurance system, such as SNAP or welfare payments, accountfor one source of difference. We exploited the available information in Annual Social Eco-nomic Supplement to the March CPS to directly assess the effect of initial unemploymentrates on receipt of transfer income. The results are shown in Figure 5 and Table 3. We finda persistent increase in the probability of receiving of SNAP benefits, but no effect for otheroutcomes, such as unemployment insurance, receipt of Earned Income Tax Credits, or wel-fare receipt (not shown). This confirms that unemployment insurance in particular does littleto smooth adverse initial conditions in the labor market for young workers, partly because oflack of eligibility and limited duration and partly because a substantial portion of the effectgoes through persistent reductions in wages.

While the point estimates of the effect on SNAP benefits appear small, they are preciselyestimated even up to 10 years after labor market entry. Relative to the average fraction ofthe sample receiving SNAP benefits, the effect is non-negligible. For a three-point rise inunemployment rates, the initial effect is about a 1.5 percentage point rise (a 15% rise withrespect to the mean in Table 1) and the cumulated effect over ten years is approximately 3.7points. Given that about 10% of individuals receive SNAP benefits, and 23% of Americansreport to have received benefits at some point, these are substantial effects. We also analyzedthe effect of initial unemployment rates on the amount of benefits received among SNAPrecipients (not shown). We find that in the first 4 to 5 years after entry, income from SNAPbenefits rose by a precisely estimated 2–3%. This has the potential to explain a substantialpart of the difference between labor earnings and household income shown in Figure 2.23

23Given these magnitudes line up well, we did not separately analyze the potential insurance stemming froma rise in spousal earnings.

17

3.2.3 Effects on Health Insurance

As health insurance receipt was chiefly tied to employment throughout most of the periodunder study, the employment effects we find can imply a loss in private health insurancecoverage. Based on the ASEC, we can analyze receipt of private health insurance, avail-ability of any health insurance, and receipt of Medicaid as separate outcomes. As expected,we see in Table 3 that exposure to a high initial unemployment rate leads to a reduction inthe incidence of private health insurance. Consistent with our findings on the reduction inemployment, the effect fades after about four to five years in the labor market. In contrast,the losses in overall access to health insurance are concentrated in the first couple of yearsafter labor market entry. The difference is explained by a persistent rise in the incidence ofreceiving Medicaid, shown in Figure 5. As shown in Table 3, this effect is significantly dif-ferent from zero for about 7 years after labor market entry. The point estimates are small buthave to be compared against the average of 10% in Medicaid receipt in our sample (see Ta-ble 1). Scaling the coefficient again by 3, suggests labor market entry in recessions increasesprobability of receiving Medicaid by about 12% relative to the baseline.

We have also analyzed the effect on family outcomes, such as marital status, childbear-ing, single parenthood, or living with parents. In contrast to other studies suggesting thatincreasing unemployment rates delay household formation, we find no effect of initial un-employment rate on marriage or childbearing. However, results shown in the Web Appendix(Section VIII) suggest there is an increase in the incidence of individuals living with theirparents. This rise only lasts for four to five years after entry, suggesting that individuals moveinto a place of their own once they have found stable employment.24

3.3 The Effect on Earnings, Employment, and Wages by Education, Race, and Gen-der

Figure 7 and Table 2 show the effects of initial unemployment rates on annual earnings andfamily income by gender and by whites versus nonwhites. The results for men and womenare qualitatively quite similar, with the exception of the effects in the first years after labormarket entry. Initially, men experience both larger losses in earnings and in family income.Turning to race, the losses in annual earnings for nonwhites are substantially larger than forwhites, especially in the first five years in the labor market. As shown in Figure 9, this dif-ference is mainly driven by greater employment losses for nonwhites. In contrast, the effect

24While these results are precisely estimated for the full sample, there was no sufficient precision to replicateit by demographic or education groups.

18

on household income is quite comparable, suggesting that social insurance mechanisms helpto buffer the bigger effects for nonwhites, something which we turn to below.

Considering Figure 9, it is remarkable how, despite some differences in the initial effect,all considered groups experienced persistent losses in hourly wages. In contrast, the tem-porary losses in employment (as measured by weeks worked last year) are somewhat moredisparate across groups. In particular, men appear to experience a more persistent reductionin employment, and nonwhites experience the largest losses. Yet, after five years in the labormarket, these effects have completely faded.

Figure 8 shows that there are substantial differences in the effect of adverse initial con-ditions by education groups. Middle educated workers—those with high school or somecollege—experience patterns comparable to that for the full sample shown in Figure 2. Incontrast, college graduates show markedly smaller and shorter-lived effects on annual earn-ings. The size of the effect is about half of that of the full sample and more similar inmagnitude to effects found for college graduates in Canada by OWH: they are somewhatsmaller than findings by Kahn (2010) for college graduates entering during the severe 1982recession. The largest effects we find are the initial losses in annual earnings experiencedby high school dropouts. These average a 5% reduction over the first three years and thenconverge in a similar fashion as the effect for those with a high school degree. The resultsare substantial losses in cumulated earnings for this group. Not surprisingly, governmenttransfers play a particularly important role in delivering a more muted impact on income forhigh school dropouts, something we return to below.

As shown in Figure 10 to an important degree, the differences in earnings losses re-sult from differences in employment losses. high school dropouts experience substantialemployment reductions, whereas college graduates essentially experience no significant em-ployment reduction. In contrast, again all groups experience reductions in hourly wageslasting at least 10 years into the career. As shown in Table 5, even for workers with at leasta college degree or with some college, the reduction is statistically significantly differentfrom zero 10 years after labor market entry. Interestingly, while reduced weekly hours isa phenomenon relevant for all demographic and education groups, there is little noticeablevariance in the effect of adverse initial labor market conditions.

19

3.4 The Effect on Other Outcomes by Education, Race, and Gender

3.4.1 Welfare Effects by Demographic Groups

Figure 11 shows the impact of initial unemployment rates on our two key variables intendedto capture the role of the social insurance system—receipt of SNAP and Medicaid—by de-mographic groups. Women have a slightly higher rate of initial receipt of SNAP, and thismay contribute to their lower income losses, but the difference is relatively small. However,we see a large difference in the propensity to receive SNAP benefits by race groups. In thefirst five years, nonwhites have an increased probability of receiving SNAP benefits of onepercentage point (Table 3). At a mean receipt of 18% in our population, that 3-point rise inunemployment rates leads to a 15% increase relative to the mean. For whites, the effect is 0.4and 0.3 points in years 1–3 and 2–5 respectively. For a 3-point recession, this implies approx-imately an 11% effect relative to the mean (8%). Interestingly, for all demographic groups,conditional on receiving SNAP benefits, the increase in the amount is relatively similar (notshown).

We also assessed the effect on other potential sources of transfer income. Although wedid not find precisely estimated effects for either men or both race groups, we found a non-negligible rise in the amount of unemployment insurance income received for women.

Overall, despite the increased probability of receiving SNAP benefits and its effect onincome, entering the labor market during slack labor markets has a significant effect onpoverty rates. Figure 13 shows that poverty rates rose persistently for the first 5 years afterlabor market entry—and in some cases even up to nine years after—for all demographicgroups, an effect that was largest for blacks. The effects are non-negligible: relative to themean poverty rates for the different demographic groups in our age ranges shown in Table 1,the effects are in the range of 10–15%.

Figure 11 also shows the rise in the incidence of receipt of Medicaid by demographicgroups. There are little discernible differences by gender, with relatively short-lived effects.The same is true for whites. In contrast, for nonwhites, the figure and table show preciselyestimated larger increases lasting 10 years into the labor market. A 3-point rise in the initialunemployment rate is predicted to trigger a rise of approximately 1.5 points initially (aneffect of about 10% relative to the mean, see Table 1) and a rise of approximately one pointthereafter.

The Medicaid results relate to the coverage by health insurance more generally. Table 3shows that there is a steep loss in private health insurance concentrated in the first 2 years in

20

the labor market for whites and both genders (with men suffering slightly larger losses) andin the first 4 years for nonwhites. Due to the increased probability of receiving Medicaid, theeffect on having any health insurance shown in the Web Appendix (Section X) is substan-tially muted for women and nonwhites and reduced for men and whites as a whole. Hence,it appears Medicaid is successful at providing a partial buffer against the temporary loss inemployer-provided health insurance.

3.4.2 Welfare Effects by Four Education Groups

Figure 12 and Table 1 show the effect of initial unemployment rates on receipt of SNAPbenefits and Medicaid by education groups. The results are quite clear. Labor market en-trants with some college or more do not appear to experience an increase in the receipt ofthese social insurance programs due to adverse initial labor market conditions. Labor marketentrants with 12 years of education experience a precisely estimated but moderate increasein the probability of obtaining these programs lasting up to 7 years after labor market en-try. In contrast, high school dropouts experience substantial increases in the probability ofreceiving SNAP that, albeit declining somewhat over time, lasts up to 15 years into the labormarket. This suggests that the effects for broader groups discussed so far are mainly drivenby responses for lower-skilled workers. The amounts received are higher initially and thendecline somewhat for college dropouts, but the changes with experience are not preciselyestimated (see the Web Appendix, Section X).

When considering unemployment insurance benefit receipt, we found relatively impre-cisely estimated effects for high school dropouts in the first couple of years after labor marketentry, but no effects for any of the other groups.

Taken together, the results on SNAP benefits and household income suggest that socialinsurance mechanisms do buffer the effect of adverse conditions at initial labor market entry.However, the insurance is imperfect. As shown in Figure 14, poverty rates rise persistentlyfor both high school graduates and high school dropouts. Given typical poverty rates in oursample of a bit over 20% (Table 1), the estimates imply a rise in poverty rates of approx-imately 15% relative to the average poverty rate during a moderate recession. It would be25% in a larger downturn with a five point rise in unemployment rates, such as the GreatRecession.

Figure 12 and Table 1 also show the effect on Medicaid. Again, as expected, there isno increase in the probability of receiving Medicaid by labor market entrants with somecollege or more. In contrast, less-educated labor market entrants see a persistent increase in

21

Medicaid receipt following an adverse labor market entry. Relative to mean Medicaid receiptfor the lowest education group (20.5%, Table 1), the effect is approximately 5–10%.

Again, these pattern are connected to health insurance availability more generally (re-sults shown in the Web Appendix, Section X). Interestingly, temporary losses in employer-provided health insurance are concentrated among high school graduates and those entrantswith some college; in contrast, neither high school dropouts nor college graduates experiencea reduction. As a result, high school graduates can partly rely on Medicaid to help bufferthe temporary loss of employer-provided health insurance, whereas those with some collegecannot.

Overall, it appears that persistent increases in the receipt of SNAP benefits and Medicaidhelp those young labor market entrants that are most affected by adverse initial labor marketconditions—nonwhites and high school dropouts, and to some degree high school graduates.The buffer appears most successful for access to any health insurance, which appears to de-cline only in the immediate years after graduation when employment losses are most severe.In contrast, the rise in transfer payments cannot prevent a temporary but long-lasting risein poverty, which reflects the large and persistent earnings losses that these less-advantagedgroups experience upon graduating during recessions.

4 Conclusion

Economists and policy makers alike have long been concerned about whether young work-ers entering the labor market during a recession suffer permanent consequences from theirinitial bad luck. While this question has been studied extensively for male college gradu-ates, less advantaged workers such as lower-educated workers, nonwhites, and women havereceived less attention. To be able to study the long-term effects for these smaller groups,we have introduced a new approach to deal with selective migration and labor market entrythat exploits information on place of birth available in the Decennial Census and the Amer-ican Community Survey. This approach allowed us to exploit cross-sectional data from theCurrent Population Survey, which spans over 4 decades, to study the effect of adverse initialconditions for multiple groups of workers. Our data did not only allow us to study workersthat are at higher risk of lasting adverse consequences. The data also allowed us to analyzewhether the adverse effects on earnings are buffered by the social insurance system for theseless-advantaged workers.

We confirm that all labor market entrants experience persistent reduction in earnings, em-ployment, and wages from entering the labor market in a recession that last at least 10 years.

22

We show these effects are substantially larger for less-advantaged workers, in particular highschool dropouts and nonwhites, but also for high school graduates. The losses in earningswe find are partly offset by increases in the receipt of SNAP benefits for the least advantagedgroups, reducing the impact on the reduction in household income. Nevertheless, our resultsimply that entering the labor market leads to persistent increases in poverty.

Overall, these findings help to complete the picture of persistent consequences of cycli-cal conditions for young workers. It becomes increasingly apparent that adverse early labormarket conditions affect all groups in the population and influence many aspects of individ-ual workers’ socio-economic outcomes. These findings highlight several important and as ofyet open questions. We know relatively little so far as to the sources of the persistent reduc-tion in employment and wages that we and others document. An important source of wagelosses for college graduates appears to be a reduction in employer quality (Oreopoulos et al.,2012). This is consistent with the fact that employment fluctuations are more pronounced athigher paying employers (e.g, Kahn and McEntarfer, 2014), leading to cyclical downgradingof labor (e.g., McLaughlin and Bils, 2001). Similar forces are likely to be present for lower-skilled labor, who are at the bottom of the job ladder. Another important question concernsthe longer-term consequences of adverse initial labor market conditions. Gibbons and Wald-man (2006), for example, hypothesize that worse occupational outcomes and human capitalaccumulation makes these workers more vulnerable to future economic shocks. In contrast,findings by Schmieder and Von Wachter (2010) suggest that below-average wages as a re-sults of adverse initial conditions may reduce the chance of future layoff. Vulnerability orresilience may also be present in long-term health outcomes. The study of these and otherquestions must be postponed until data with additional information in career outcomes andlonger time ranges are available.

23

References

Altonji, Joseph G., Lisa B. Kahn, and Jamin D. Speer, “Cashier or consultant? Entrylabor market conditions, field of study, and career success,” Journal of Labor Economics,2016, 34 (S1), S361–S401.

Autor, David H, David Dorn, Gordon H Hanson, Jae Song et al., “Trade adjustment:Worker-level evidence,” The Quarterly Journal of Economics, 2014, 129 (4), 1799–1860.

Blanchflower, David G. and Andrew J. Oswald, “Estimating a Wage Curve for Britain1973-1990,” Economic Journal, 1994, 104 (426), 1025–1043.

Card, David and Alan B Krueger, “Does school quality matter? Returns to education andthe characteristics of public schools in the United States,” Journal of Political Economy,1992, 100 (1), 1–40.

Chetty, Raj and Nathaniel Hendren, “The impacts of neighborhoods on intergenerationalmobility I: Childhood exposure effects,” The Quarterly Journal of Economics, 2018, 133(3), 1107–1162.

, , and Lawrence F Katz, “The effects of exposure to better neighborhoods on children:New evidence from the Moving to Opportunity experiment,” The American EconomicReview, 2016, 106 (4), 855–902.

Currie, Janet and Hannes Schwandt, “Short-and long-term effects of unemployment onfertility,” Proceedings of the National Academy of Sciences, 2014, 111 (41), 14734–14739.

Cutler, David M, Wei Huang, and Adriana Lleras-Muney, “When does education matter?The protective effect of education for cohorts graduating in bad times,” Social Science &Medicine, 2015, 127, 63–73.

Deaton, Angus, “Panel data from time series of cross-sections,” Journal of Econometrics,1985, 30 (1-2), 109–126.

Elsby, Michael WL, Donggyun Shin, and Gary Solon, “Wage adjustment in the Great Re-cession and other downturns: Evidence from the United States and Great Britain,” Journalof Labor Economics, 2016, 34 (S1), S249–S291.

24

Gardecki, Rosella and David Neumark, “Order from chaos? The effects of early labormarket experiences on adult labor market outcomes,” Industrial and Labor Relations Re-view, 1998, 51 (2), 299–322.

Gibbons, Robert and Michael Waldman, “Enriching a theory of wage and promotion dy-namics inside firms,” Journal of Labor Economics, 2006, 24 (1), 59–107.

Giuliano, Paola and Antonio Spilimbergo, “Growing up in a Recession,” The Review ofEconomic Studies, 2014, 81 (2), 787–817.

Hoynes, Hilary, Douglas L. Miller, and Jessamyn Schaller, “Who suffers during reces-sions?,” Journal of Economic Perspectives, 2012, 26 (3), 27–48.

Kahn, Lisa B, “The long-term labor market consequences of graduating from college in abad economy,” Labour Economics, 2010, 17 (2), 303–316.

and Erika McEntarfer, “Employment cyclicality and firm quality,” 2014. NBER Work-ing Paper 20698, National Bureau of Economic Research, Cambridge, MA.

Maclean, Johanna Catherine, “The health effects of leaving school in a bad economy,”Journal of Health Economics, 2013, 32 (5), 951–964.

McLaughlin, Kenneth J and Mark Bils, “Interindustry mobility and the cyclical upgradingof labor,” Journal of Labor Economics, 2001, 19 (1), 94–135.

Murphy, Kevin M and Finis Welch, “Empirical age-earnings profiles,” Journal of Laboreconomics, 1990, 8 (2), 202–229.

Oreopoulos, Philip, Till Von Wachter, and Andrew Heisz, “The short-and long-term ca-reer effects of graduating in a recession,” American Economic Journal: Applied Eco-nomics, 2012, 4 (1), 1–29.

Oyer, Paul, “Initial labor market conditions and long-term outcomes for economists,” TheJournal of Economic Perspectives, 2006, 20 (3), 143–160.

, “The making of an investment banker: Stock market shocks, career choice, and lifetimeincome,” The Journal of Finance, 2008, 63 (6), 2601–2628.

Pei, Zhuan, Jörn-Steffen Pischke, and Hannes Schwandt, “Poorly measured confoundersare more useful on the left than on the right,” Journal of Business & Economic Statistics,2018, 0 (0), 1–12.

25

Rothstein, Jesse and Robert G Valletta, “Scraping by: Income and program participationafter the loss of extended unemployment benefits,” Journal of Policy Analysis and Man-agement, 2017, 36 (4), 880–908.

Schmieder, Johannes F and Till Von Wachter, “Does wage persistence matter for em-ployment fluctuations? Evidence from displaced workers,” American Economic Journal:Applied Economics, 2010, 2 (3), 1–21.

Schwandt, Hannes and Till Von Wachter, “Mortality Profiles of Unlucky Cohorts: Effectsof Entering the Labor Market in a Recession on Longevity,” 2018. Working paper, UCLA.

Speer, Jamin D., “Wages, hours, and the school-to-work transition: The consequences ofleaving school in a recession for less-educated men,” The BE Journal of Economic Analy-sis & Policy, 2016, 16 (1), 97–124.

Sullivan, Daniel and Till Von Wachter, “Job displacement and mortality: An analysis usingadministrative data,” The Quarterly Journal of Economics, 2009, 124 (3), 1265–1306.

Topel, Robert H and Michael P Ward, “Job mobility and the careers of young men,” TheQuarterly Journal of Economics, 1992, 107 (2), 439–479.

Wachter, Till Von, “Job displacements in recessions: An overview of long-term conse-quences and policy options,” in “Reconnecting to Work: Policies to Mitigate Long-TermUnemployment and Its Consequences,” ed. L.D. Appelbaum. Kalamazoo: W.E. UpjohnInstitute for Employment Research, 2012.

and Stefan Bender, “In the right place at the wrong time: The role of firms and luck inyoung workers’ careers,” The American Economic Review, 2006, 96 (5), 1679–1705.

26

Figure 1: Mincerian and Double-Weighted graduation year unemployment rates acrossfour states

Notes: The Mincerian graduation year unemployment rate (blue circles) refersto the unemployment rate in the current state of residence at the “Mincerian”year of graduation (the sum of the year of birth, plus 6, plus the years of re-ported education). The double-weighted graduation year unemployment rate(red line) refers to the average unemployment rate across cohorts’ typical grad-uation ages and across the states to which cohorts typically migrate beforegraduation. See sections 2.2 and 2.3 for further details. There is only onedouble-weighted unemployment rate for each birth cohort, but several Min-cerian rates given different graduation times and migration to different states.Rates above 13 and below 3 are omitted in order to preserve a transparent scale.

27

Figure 2: Effect of State Unemployment Rate at Labor Market Entry on Log AnnualEarnings and Household Income for Full Sample

−.0

4−

.03

−.0

2−

.01

0E

ffect

on lo

g e

arn

ings/

inco

me

1 3 5 7 9 11 13 15Years since Graduation

Annual Earnings Annual Household Income

Notes: Results are based on the Mincerian specification (equation 2), usingdata from the ASEC Supplement to CPS from 1976 to 2016.

28

Figure 3: Adjusting for Migration and Graduation Timing in Decennial Census andAmerican Community Survey Data

−.0

4−

.02

0.0

2E

ffect

on lo

g e

arn

ings

1 [19] 5 [23] 9 [27] 13 [31]

Years since Graduation [Age]

CPS, Mincerian (baseline) Census, Mincerian

Census, using state of birth Census, Double−Weighted [by age]

Notes: The CPS baseline results are based on the ASEC Supplement to CPSfrom 1976 to 2016. The Census results are based on the 1980/1990/2000 De-cennial Censuses and the ACS from 2001 to 2015. The Mincerian graduationyear unemployment rate refers to the unemployment rate in the current state ofresidence at the “Mincerian” year of graduation (the sum of the year of birth,plus 6, plus the years of reported education). The double-weighted graduationyear unemployment rate refers to the average unemployment rate across co-horts’ typical graduation ages and across the states to which cohorts typicallymigrate before graduation. See sections 2.2 and 2.3 for further details.

29

Figure 4: Using the Double-Weighted Average Graduation Year Unemployment Rateas Instrumental Variable for State Unemployment Rate at Labor Market Entry

−.0

6−

.04

−.0

20

.02

Effect

on lo

g e

arn

ings

1 [19] 5 [23] 9 [27] 13 [31]

Years since Graduation [Age]

CPS, Mincerian (baseline) Census, Mincerian

Census, Double−Weighted Double−Weighted as IV for baseline

Notes: The CPS baseline results are based on the ASEC Supplement to CPSfrom 1976 to 2016. The Census results are based on the 1980/1990/2000 Cen-suses and the ACS from 2001 to 2015. The Mincerian graduation year unem-ployment rate refers to the unemployment rate in the current state of residenceat the “Mincerian” year of graduation (the sum of the year of birth, plus 6, plusthe years of reported education). The double-weighted graduation year unem-ployment rate refers to the average unemployment rate across cohorts’ typicalgraduation ages and across the states to which cohorts typically migrate beforegraduation. See sections 2.2 and 2.3 for further details.

30

Figure 5: Effect of State Unemployment Rate at Labor Market Entry on Employmentand Wages

−.0

2−

.015

−.0

1−

.005

0E

ffect

on lo

g o

utc

om

e


Hourly Earnings Usual Hours Worked

Weeks Employed


31

Figure 6: Effect of State Unemployment Rate at Labor Market Entry on SNAP Benefitsand Medicaid for Full Sample

−.0

04

−.0

02

0.0

02

.004

.006

Effect

on fra

ctio

n r

ece

ivin

g w

elfa

re


Medicaid Foodstamps


32

Figure 7: Effect of State Unemployment Rate at Labor Market Entry on Earnings andIncome by Demographic Groups

−.0

6−.0

4−.0

20

.02

Effect

on lo

g in

com

e


Male

−.0

6−.0

4−.0

20

.02


Female

−.0

6−.0

40

.02

−.0

2


White−

.06−

.04−

.02

0.0

2


Non−white

Annual Earnings Annual Household Income


33

Figure 8: Effect of State Unemployment Rate at Labor Market Entry on on Earningsand Income by Education Groups

−.0

6−

.04

−.0

20

.02

Effect

on lo

g in

com

e


Figure 9: Effect of State Unemployment Rate at Labor Market Entry on Employmentand Wages by Demographic Group

−.0

3−

.02−

.01

0E

ffect

on lo

g o

utc

om

e


Male

−.0

3−

.02−

.01

0


Female

−.0

3−

.02−

.01

0


White

−.0

3−

.02−

.01

0


Non−white

Hourly Earnings Usual Hours Worked

Weeks Employed


35

Figure 10: Effect of State Unemployment Rate at Labor Market Entry on Employmentand Wages by Education Groups

−.0

4−

.02

0.0

2E

ffect

on lo

g o

utc

om

e


Figure 11: Effect of State Unemployment Rate at Labor Market Entry on SNAP Ben-efits and Medicaid by Demographic Groups

−.0

05

0.0

05

.01

.01

5E

ffect

on w

elfa

re


Male

−.0

05

0.0

05

.01

.01

5


Female

−.0

05

0.0

05

.01

.01

5


White−

.00

50

.00

5.0

1.0

15


Non−white

Medicaid Foodstamps

Notes: Results are based on the ASEC Supplement to CPS from 1976 to 2016.

37

Figure 12: Effect of State Unemployment Rate at Labor Market Entry on SNAP Ben-efits and Medicaid by Education Groups

−.0

10

.01

.02

Effect

on w

elfa

re


Figure 13: Effect of State Unemployment Rate at Labor Market Entry on Incidence ofPoverty by Demographic Groups

−.0

05

0.0

05

.01

Effect

on fra

ctio

n p

oor


Male Female

White Non−white


39

Figure 14: Effect of State Unemployment Rate at Labor Market Entry on Incidence ofPoverty by Demographic Groups

−.0

05

0.0

05

.01

.015

Effect

on fra

ctio

n p

oor


Less than 12 Equal to 12

Between 13 and 15 Over 15


40

41

Full Sample

Men Women White Non-White

Less than 12 Years

of Schooling

12 Years of

Schooling

13-15 Years of

Schooling

16 or More

Years of Schooling

Average Annual Earnings Last Year 18,343$ 22,125$ 14,492$ 19,115$ 15,349$ 4,143$ 13,999$ 17,065$ 34,750$

Aveerage Hourly Earnings Last Year 13$ 14$ 12$ 13$ 13$ 8$ 10$ 12$ 20$

Average Household Income Last Year 57,415$ 59,481$ 55,311$ 59,764$ 48,313$ 46,995$ 46,951$ 56,744$ 77,701$

Average Weeks Worked Last Year 34.54 37.21 31.81 35.73 29.90 16.93 35.32 37.50 42.91

Average Usual Weekly Hours in Survey Week 37.91 40.01 35.51 38.01 37.48 30.45 38.27 37.27 41.17

Employed in Survey Week 0.69 0.73 0.65 0.71 0.60 0.38 0.70 0.73 0.85Fraction Receiving SNAP (Food Stamps) 0.10 0.08 0.12 0.08 0.18 0.20 0.13 0.07 0.02Average Value SNAP Benefits 2,306$ 2,209$ 2,369$ 2,135$ 2,581$ 2,591$ 2,200$ 2,098$ 1,783$

Fraction with Any Health Insurance 0.75 0.72 0.79 0.77 0.71 0.68 0.67 0.76 0.88

Fraction with Private Insurance 0.67 0.67 0.67 0.70 0.56 0.49 0.56 0.70 0.87

Fraction Receiving Medicaid 0.10 0.07 0.13 0.08 0.17 0.21 0.12 0.08 0.02

Table 1: Sample Statistics for Cohorts of Workers Entering the Labor Market from 1976 to 2015 with One to 15 Years of Potential Labor Market Experience

Notes: Data from ASEC Supplement Current Population Survey. Potential labor market experience is defined as age minus years of completed schooling minus 6. Year of labor market entry is the implied calendar year after the year of completion of highest level of schooling starting from age 6. All dollar values are deflated by the CPI with base year 2000. See text for further details.

42

VariableExperience

GroupFull Sample Men Women White Non-White

1-3 -0.0380 -0.0428 -0.0324 -0.0375 -0.0436(0.0023) (0.0031) (0.0029) (0.0024) (0.0060)

4-5 -0.0250 -0.0263 -0.0223 -0.0230 -0.0401(0.0024) (0.0028) (0.0033) (0.0025) (0.0057)

6-7 -0.0117 -0.0118 -0.0101 -0.0126 -0.0105(0.0023) (0.0030) (0.0031) (0.0025) (0.0056)

8-10 -0.0090 -0.0094 -0.0084 -0.0105 -0.0087(0.0022) (0.0026) (0.0032) (0.0022) (0.0057)

1-3 -0.0178 -0.0211 -0.0144 -0.0176 -0.0171(0.0017) (0.0020) (0.0021) (0.0018) (0.0034)

4-5 -0.0118 -0.0116 -0.0115 -0.0124 -0.0128(0.0017) (0.0022) (0.0022) (0.0018) (0.0044)

6-7 -0.0076 -0.0078 -0.0070 -0.0078 -0.0110(0.0017) (0.0022) (0.0021) (0.0017) (0.0044)

8-10 -0.0044 -0.0037 -0.0045 -0.0049 -0.0084(0.0017) (0.0020) (0.0023) (0.0016) (0.0045)

1-3 -0.0139 -0.0154 -0.0125 -0.0149 -0.0138(0.0013) (0.0018) (0.0017) (0.0014) (0.0033)

4-5 -0.0127 -0.0143 -0.0101 -0.0123 -0.0177(0.0015) (0.0020) (0.0020) (0.0016) (0.0038)

6-7 -0.0074 -0.0075 -0.0067 -0.0083 -0.0058(0.0014) (0.0019) (0.0017) (0.0015) (0.0036)

8-10 -0.0060 -0.0063 -0.0065 -0.0070 -0.0050(0.0013) (0.0017) (0.0019) (0.0014) (0.0036)

1-3 -0.0150 -0.0174 -0.0113 -0.0129 -0.0230(0.0014) (0.0018) (0.0020) (0.0014) (0.0038)

4-5 -0.0063 -0.0067 -0.0055 -0.0050 -0.0147(0.0013) (0.0016) (0.0020) (0.0013) (0.0034)

6-7 -0.0017 -0.0026 -0.0003 -0.0012 -0.0035(0.0013) (0.0016) (0.0019) (0.0013) (0.0032)

8-10 -0.0008 -0.0008 -0.0004 -0.0008 -0.0029(0.0012) (0.0014) (0.0018) (0.0012) (0.0030)

Notes: Data from ASEC Supplement Current Population Survey. Regressions control for fixed effects for potential labor market experience, calendar year, year of graduation, state of current residence, and three education groups.

Table 2: Effects of State Unemployment Rates at Labor Market Entry on Earnings, Income, Wages, and Employment for Cohorts Entering in 1976 to 2015 for the Full sample, by Gender, and by Race

Log Annual Earnings

Log Household Income

Log Hourly Wages

Log Weeks Worked

43

VariableExperience

GroupFull Sample Men Women White Non-White

1-3 0.0053 0.0050 0.0055 0.0038 0.0111(0.0005) (0.0006) (0.0007) (0.0005) (0.0016)

4-5 0.0038 0.0026 0.0049 0.0027 0.0093(0.0006) (0.0007) (0.0008) (0.0006) (0.0017)

6-7 0.0022 0.0022 0.0020 0.0017 0.0062(0.0006) (0.0008) (0.0008) (0.0006) (0.0016)

8-10 0.0011 0.0003 0.0015 0.0006 0.0056(0.0006) (0.0007) (0.0008) (0.0005) (0.0017)

1-3 0.0056 0.0065 0.0048 0.0046 0.0096(0.0006) (0.0007) (0.0008) (0.0006) (0.0014)

4-5 0.0033 0.0031 0.0032 0.0029 0.0062(0.0007) (0.0009) (0.0009) (0.0007) (0.0018)

6-7 0.0013 0.0012 0.0011 0.0015 0.0023(0.0006) (0.0008) (0.0009) (0.0006) (0.0018)

8-10 0.0014 0.0011 0.0013 0.0012 0.0047(0.0006) (0.0007) (0.0008) (0.0006) (0.0016)

1-3 0.0041 0.0037 0.0044 0.0036 0.0054(0.0006) (0.0006) (0.0007) (0.0006) (0.0015)

4-5 0.0012 0.0016 0.0005 0.0009 0.0030(0.0007) (0.0008) (0.0009) (0.0007) (0.0018)

6-7 0.0014 0.0006 0.0018 0.0014 0.0027(0.0007) (0.0007) (0.0009) (0.0007) (0.0017)

8-10 0.0009 -0.0002 0.0015 0.0008 0.0035(0.0006) (0.0006) (0.0009) (0.0006) (0.0017)

1-3 -0.0085 -0.0096 -0.0074 -0.0079 -0.0102(0.0009) (0.0012) (0.0012) (0.0010) (0.0019)

4-5 -0.0019 -0.0016 -0.0023 -0.0019 -0.0046(0.0012) (0.0016) (0.0014) (0.0012) (0.0025)

6-7 -0.0008 -0.0016 0.0000 -0.0011 -0.0021(0.0010) (0.0014) (0.0012) (0.0011) (0.0020)

8-10 0.0006 0.0005 0.0007 0.0002 -0.0011(0.0009) (0.0012) (0.0012) (0.0009) (0.0019)

Notes: Data from ASEC Supplement Current Population Survey. Regressions control for

Documents

DICION PAPER ERIE - IZA Institute of Labor Economicsftp.iza.org/dp11926.pdfIA DP No. 11926 OCTOBER 201 Unlucky Cohorts: Estimating the Long-term Effects of Entering the Labor Market